9:16, 16:9 or Square? Choose a Video Aspect Ratio
Updated: Nov 5
Choose a video’s aspect ratio for the place people will watch it and the composition the message needs. Vertical 9:16 suits a tall viewing area, landscape 16:9 suits a wide one, and square 1:1 gives equal width and height. Check the destination’s current requirements before generating: changing shape later can remove important parts of the shot.
Aspect ratio describes shape, not image quality
The numbers express width relative to height. A 1280 × 720 frame is 16:9; a 720 × 1280 frame is 9:16; a 1080 × 1080 frame is 1:1. A vertical file is not automatically a rotated landscape file. Turning a finished frame sideways also turns the subject sideways.
Resolution describes pixel dimensions. Two files can have the same aspect ratio at different resolutions, and two files with the same total number of pixels can have different shapes. Decide the frame shape first, then select an appropriate supported resolution and delivery encoding.
Use 16:9 when width carries the information
Landscape helps when the audience needs to see a room, a broad view, two people beside each other or a presentation with horizontal graphics. The extra width can preserve context around the subject. It may also suit an embedded player on a desktop page.
For example, a real estate teaser showing the relationship between a kitchen and living area needs that spatial context. A very tight vertical crop may make the room difficult to understand. Plan a separate close-up for a small detail rather than expecting one frame to explain everything.
YouTube documents 16:9 as its standard desktop aspect ratio while its player adapts to other shapes. That is a reason to consider the viewing surface, not a rule that every YouTube upload must be landscape. Review the format intended for your particular placement.
Use 9:16 when a tall composition helps
Vertical composition can work well for a standing person, a tall object or a single focused subject viewed on a phone in portrait orientation. Leave breathing room around the important content. A product that already touches both edges of the source photograph gives you little flexibility for movement or later overlays.
Plan where captions, a heading or a call to action will go in the edit. Platform interfaces can cover different parts of a video, and those interfaces change. There is no single safe margin that guarantees visibility in every app. Preview the actual destination and keep essential text away from likely controls.
What are the dimensions of a 9:16 video?
A 9:16 export can be 1080 × 1920 or 720 × 1280 pixels. The matching 16:9 examples reverse the dimensions. When adapting a landscape clip, preview where the hands, subtitles, and product fall in the narrower vertical frame.
Frame each version around its own focal point
Mark the subject and the text that must be visible in each destination. In a wide shot of a demonstration, the vertical version may need a separate crop around the hands while the landscape version keeps the full setup.
Use 1:1 when the destination and composition support it
A square frame can be useful for a compact product composition or a placement designed around a square media area. It is a choice for a particular layout, not a compromise that automatically performs best everywhere. A wide group may be awkward in a square, while a centered object may fit comfortably.
AI LEMNISCATA currently offers vertical 9:16 and landscape 16:9 in its video workflow. A square delivery may require reframing in an external editor. Choose source composition with that crop in mind and inspect the result; do not assume there is a native square generation option in the current workspace.
Cropping and fitting produce different results
Cropping fills the new frame by removing material outside it. Fitting keeps the whole original frame and leaves unused space around it, which you may treat with a deliberate background in an editor. Stretching changes the subject’s proportions and is usually inappropriate for an accurate product or person.
Consider a 1920 × 1080 landscape image cropped to 9:16 while preserving its height. The retained width is about 608 pixels: 1080 × 9 ÷ 16. That is roughly 32 percent of the original width. A centered product may survive the crop, while a second object near the edge disappears. This is geometry, not an AI failure.
Reframing lets you choose which part remains visible over time. Review the whole moving shot: a face may fit at the beginning and leave the crop halfway through. Avoid sudden automatic crop shifts that distract from the message.
Plan two versions before you generate
If you need both landscape and vertical deliverables, list the information that must remain visible in each. Prepare a wide composition for context and a tall composition for the main subject. A shared central subject helps, but one source will not always support both equally well.
In the AI LEMNISCATA video workflow, select the orientation in the first step and upload reference photos before building the plan. Check the transferred plan in the second step before checkout. Keep text overlays for a suitable editor when precise positioning matters.
Final checks and common questions
Watch the entire export in its intended orientation. Check faces, hands, product edges, captions and the final action. Upload a review version to the destination when possible and inspect its processed display, rather than judging only inside an editor.
Should I add black bars myself?
Do not add them automatically. YouTube advises against baking padding into the video because its player can adapt to the source aspect ratio. Other destinations may have different requirements. Follow the actual specification and use a deliberate fitted layout only when it serves the project.
Can I convert landscape to vertical without losing anything?
You can preserve the whole picture by fitting it inside the tall frame, but it will not fill that frame without surrounding space. Filling the frame usually requires cropping or generating additional image area. Generated surroundings need their own accuracy review.
Does 9:16 mean 1080 × 1920 specifically?
No. That is one resolution with a 9:16 shape. A 720 × 1280 file has the same aspect ratio at different pixel dimensions. Check both the ratio and the destination’s resolution requirements.
Further planning
Use vertical reference-photo composition to prepare the source, cropping versus camera movement to choose the right change, and the 720p decision guide to assess output resolution.
Reference
YouTube: video resolution and aspect ratios documents its player behavior and example resolutions. Check current placement requirements before delivery.
If the frame shape is correct but the subject plays sideways, follow How to Fix a Sideways Video Without Stretching the Picture.